Download SnackBot

SnackBot Sensor Report
Chris Shepherd
EEL 5666
7/7/05
1
Table of Contents
Behaviors and Sensors………………………………………………………….………..3
Bump Switches.……..……………….……………………………………………….….3
IR Range Finder………….……………………………………………………………...4
Line Follower……………...…………………………………………………………….4
Color Detector…………………………..……………………………………………….5
2
Behaviors and Sensors
SnackBot has four main behaviors. These are:
1. Avoid Obstacles
2. Follow a Line
3. Detect Color
4. Pick up and Object
Once SnackBot has received a command for a snack or drink he will follow a line
to the kitchen. Along the way he will use his obstacle avoidance so as to not trip
anybody who may be up and walking around the house. Once in the kitchen SnackBot
will make his way to the pantry or refrigerator and select your item of choice by its
packaging color. It will pick up your object and return it to you at the couch.
To accomplish these behaviors SnackBot will need bump switches, an IR range
finder, photo reflectors, and a special color detector.
Bump Switches
The Bump switches are attached to a bumper that wraps half-way around
SnackBot. This allows the use of only four switches to detect bumps 180 degrees around
the robot. One side of each switch is connected to a general I/O port pin which is pulled
up internally. The other side is connected to ground. Thus when the switch is not closed
the port pin reads a logic 1. And when the switch is closed the pin reads a logic 0. The
program written uses these values to determine if it has run into an object, and if it has,
on which side of the robot the object is. This is then used to change direction away from
the object.
3
IR Range Finder
The IR range finder is a Sharp GP2D12. It detects the distance to the nearest
object and returns that distance as an analog voltage. This output is connected to an A2D
capable port on the microprocessor. Using software based on some sample code from
bdmicro.com I am able to convert this analog voltage to a digital number. To help avoid
errors due to inaccurate data I take the average of 5 readings and use that as my sensor
value. I do not know the threshold value that my program uses to determine if an object
is close in front of it because this value changes every time the robot is run. This is
because I wrote a calibration routine to determine this value each time the robot is
powered up. This calibration helps reduce errors due to different lighting conditions.
Line Follower
The line follower is made up of 4 Hamamatsu photo reflectors. I have not made
the board that will hold all four photo reflectors, but I have tested the photo reflectors
individually. Although the photo reflectors return an analog value like the IR range
finder, no A2D conversion is necessary. This is because I am only interested in 2 states
when following the line. I simply need to know whether the robot is on the line, or
whether it’s off the line. The line will be a high contrast line: a black line on a white
surface. Simply needing to know whether it’s on the line or not means only needing to be
able to differentiate between black and white. The photo reflectors return a value of
approximately 5 volts if placed over a black surface, and a value of approximately 0 volts
if over a white surface. Since these are the logic levels of the microprocessor there is no
4
need for a conversion. The photo reflectors will be set up in the circuit seen below. The
output will be passed to a general I/O port pin and read in through software.
Color Detector
The color sensor consists of a red, blue, and green LED, and a single CDS cell.
The theory behind this sensor is that different objects reflect light of different
wavelengths. This is what causes an object to appear the color that it is. For example, a
red object appears red in white light because it reflects the wavelengths of light that
appear red, and absorbs all the other wavelengths. So, if the same object were placed in
green light, it would still absorb all non-red wavelengths, but since this is green light and
there are no red wavelengths to reflect, the object would not be able to reflect any light. I
use this concept and measure the amount of light reflected under different lighting
conditions to determine a color.
This sensor is made up of a CDS cell and three LEDs. A CDS cell varies in
resistance based upon the amount of light it detects. The color sensor will use the LEDs
5
to shine only red, blue, or green light at a time. A red object will reflect all red light,
causing the CDS cell to have a very different resistance than if a blue or green object,
which will absorb red light, were present. This difference in resistance will be used in a
voltage divider circuit to create different voltages to signal SnackBot if the object being
tested is red, blue, or green.
The table below shows the different resistance values of the CDS cell with several
objects under differing lighting conditions. These are most likely not the final threshold
values I will use in my software as the resistance values are likely to change once the
sensor is attached to the robot.
Object
Red LED
Blue LED
Green LED
None
130 k
240 k
300 k
Red
5.3 k
160 k
130 k
Blue
11 k
55 k
190 k
Green
10 k
130 k
80 k
With these values an easy algorithm to detect color has been determined.
Blue object if: resistance is less than 100 kOhms under blue light
Green object if: resistance is less than 100 kOhms under green light
Red object if: resistance is greater than 100 kOhms under blue and green light and
resistance is less than 10 kOhms under red light
6